# How to Value a Stock with Earnings Power Value (EPV)

Written by

Jae Jun

What You Will Learn

• How to use the EPV and improve your investing
• A step by step walkthrough using MSFT as an example
• How EPV stacks up against DCF and the Ben Graham formula

### How to Value Stocks Series

For other posts in the series, follow the links below.

#### Valuing a Stock with Earnings Power Value (EPV)

I’ve been focusing a lot of my time for the past 3 weeks dissecting and reverse engineering Bruce Greenwald’s earnings power value EPV method and it’s time I performed a stock value calculation based on EPV.

Microsoft (MSFT) will serve as a fine example since you know the history of the company and what it does. I’ll then compare the EPV valuation price with a DCF value calculation and Benjamin Graham’s formula. I’ll try to add as much information for those that haven’t read Greenwald’s EPV book.

(A big thanks to Jim Hodges for providing excellent feedback and discussions on the addition to the investment spreadsheet. Note: This is a long tutorial/analysis.)

#### Earnings Power Value Technique

The valuation technique of earnings power value requires the investor to consider 3 things.

1. The value of assets a competitor will be required to have in order to achieve the same market value of the incumbent company in the industry.
2. Earnings power value calculated based on current financial status where the resulting intrinsic value ignores business cycles.
3. Whether growth is a factor. Growth is usually ignored in this valuation technique though, so I won’t be going into the growth aspect.

#### Reproducing the Assets of Microsoft

In respect to no.1 let’s say a company is currently in the business of selling inkjets for printers. The company has a market value of \$1m but when we analyze the assets we find out the company assets are worth around \$500k. This means that if I was able to reproduce the same assets for \$500k, I should be able to create a company that is valued at \$1m in the market.

Applying this idea to Microsoft, the first step is to adjust the balance sheet. My initial mistake with this step was to look at discounting most of the items such as inventory and intangibles, but that isn’t the purpose of the asset reproduction. We are trying to get to a figure that a competitor will have to realistically pay up in order to enter the market. This is why I’ve left mostly everything in tact because for another software company to compete with Microsoft’s Windows operating system, office suites and increasing Bing market share, a company will have to fork out money for intangibles and goodwill in order to acquire technology and other company’s intelligence.

The only adjustment I made was to reduce goodwill by 50% because I know that Microsoft haven’t made the best of decisions regarding prices paid for acquisitions and the current balance sheet does not reflect Microsoft’s stake in Yahoo yet.

earnings power value asset valuation

So the new adjusted asset value is \$68,916 million.

Step two. The next step is to realize that for a company like Microsoft, there is going to be value added to the company coming from its marketing and R&D. No competitor will be able to compete if they do not try to spend money to increase brand awareness and on R&D efforts.

Some companies will spend very little for both aspects and it can be ignored but to ignore for MSFT would be a mistake.

The amount of marketing added back to the asset value above is calculated by

Taking the average of marketing, general, administration (MGA) as a % of sales for the most recent 5 years and then multiple the % to the current sales figure.

We do the same thing with R&D except for a company such as MSFT, R&D will be a very valuation asset.

The book describes a few ways of going about it but I just simply

take the sum of the past 3 years of R&D and then take 80%.

The number that I use can be understating the reproduction cost  as the competitor could need more than 5 years of R&D in order to be competitive but that’s up to you to consider.

EPV Asset Valuation for MSFT

Finally, add the marketing and R&D value of \$17,631m and \$17,495m respectively to \$68,916m we got in step one to get \$104,403m.

Something else to remember is that although off balance sheet liabilities are liabilities, a significant part of that will also have to be reproduced by a new entrant in order to start business.

E.g. VVTV may be just another home shopping company but for a new competitor to enter the market, they will have to spend money on carriage licenses and other network equipment and licenses that will sometimes not appear on the balance sheet. It may be a liability when valuing the business as a standalone, but when considering what a competitor will have to pay, it should be included.

The final step to calculate the net reproduction cost is to subtract non interest bearing debt and the cash not required to run the business.

Non interest bearing debt is really spontaneous liabilities. Total liabilities isn’t used because it could also include items that are not related to the business such as liabilities for damages, something a new entrant won’t have to pay for.

The formula I use for “cash not required for operations” is

cash and cash equivalents – 2% of sales

Greenwald mentions a couple of times that in general, 1% of sales is the amount required for a company to continue running operations. I’ve used 2% for good measure for this item.

Subtract non interesting bearing debt and excess cash from the \$104,403 figure to get the net reproduction cost of \$41,181m which is equal to \$4.63 per share.

EPV Asset Valuation MSFT IMG2

This means that a new potential competitor will have to spend around \$41billion in order to compete with Microsoft.

But you don’t just finish off with a net reproduction cost, you now have to calculate the EPV and compare it to the reproduction cost to determine the company’s competitive advantage.

#### Earnings Power Value Calculation

In the asset valuation section, you had to make adjustments to the balance sheet and now we’ll have to make adjustments to the income statement to come up with an adjusted income.

The concept is very similar. Start off with EBIT, and start working through items and decide whether to add it back or ignore it.

(click to enlarge)

Earnings Adjustment Using EBIT

First step. Start off with operating earnings, i.e. EBIT. Find out if there are any one time charges and add it back. I haven’t bothered to go back to the reports with this analysis.

In the image above, you see that the EBIT margin from 2005 to 2008 is around 40% while 2009 is at 33.9%. You can adjust this to say that Microsoft is likely to have an EBIT margin of 40% based on regular business conditions and enter 40% into the yellow box on the right side under user input, or you can just continue along with the current numbers.

I’ll just continue.

Adjusting the Earnings for EPV

Second step is to add a certain percentage of SG&A and R&D back to earnings. I prefer to keep things simply by adding back 25% for both. This now leaves you with what we call Adjusted EBIT. Compare the adjusted EBIT percentage margin to the stated EBIT margin.

Third step is to apply a tax rate to the adjusted EBIT. Since EBIT is earnings before interest and taxes, if we pay taxes on EBIT, it now simply becomes earnings. Which is represented by the Adjusted Earnings After Tax line in the image above.

Fourth step you add back in a certain amount of depreciation and amortization. The best thing is to be familiar with the business and industry to accurately assess the equipment needed and how fast it loses value etc. The easy way would be to add back 20% of D&A as I do.

Get instant access to value investing spreadsheets, how to's and stock ideas.

When you do all those steps, you finally come up with an Income as Adjusted number. What I do, along with all of the other valuation techniques, I smooth out the data by taking multiple year snapshots and then taking the median of these timeframes. This way I come up with a normalized adjusted income to ignore business cycles and the occasional overly bad or good year.

Earnings Power Value Adjusted Income and Growth

So with the normalized adjusted income you subtract maintenance capital expenditures and divide by the discount rate. I used a simple 9% in this example. I don’t bother with WACC as it is seriously flawed due to its dependence on beta.

The result is the EPV, which is the value of the company based on current earnings and ignoring growth. But there is one last step.

Lastly, add to the EPV value Cash-debt because operating earnings ignore the interest on cash balances so you have to add the surplus cash to the EPV.

EPV Final Caluation

So the EPV of Microsoft is \$24.36 and the reproduction value is \$4.63.

What does this mean?

It means that the \$19.73 difference is the competitive advantage enjoyed by Microsoft. Refer to slide 18 of Greenwald’s EPV lecture slide.

How does this valuation fair with discounted cash flow calculation and Benjamin Graham’s formula?

#### Discounted Cash Flow Calculation

DCF Value to Price Graph for MSFT

According to the DCF calculator, using the same 9% discount rate with the along with the default calculation of 9.2% growth gives an intrinsic value of \$27.91.

Remember that I haven’t bothered making any changes to any of the inputs throughout this entire analysis. Everything is just based off automatic calculations so you could fine tune your results if desired.

#### Benjamin Graham Formula Valuation

Ben Graham Formula Valuation

In this case, the normlized earnings (EPS) growth by Microsoft over the past 10 years has been at 13.8% with normalized earnings at \$1.01. Applying Benjamin Graham’s formula, the intrinsic value comes out to \$30.04.

#### Stock Value Calculation Comparison

Stock Valuation Comparisons

Reproduction value of \$4.63 shows that Microsoft has a big competitive advantage.

EPV of \$24.63 is the stock value based on current financial results.

DCF valuation of \$27.91

Benjamin Graham valuation of \$30.04

Seems like all three match up well and I can confidently say that Microsoft is fairly valued between \$24 and \$30. The company is currently trading at \$25.26 which places it smack into the fair value range.

#### Final Thoughts

EPV is a great valuation technique but shouldn’t be used just on its own. Compare it with other valuation methods and it’s a great addition to any investors knowledge base.

And just before people start asking about when they can receive the updated spreadsheet, all premium buyers will be receiving it in October. If you don’t have it yet, get it before October when I raise prices. More on this later.

View the pdf below or download the file from scribd.
Microsoft MSFT Stock Valuation Calculator Spreadsheet

Get my customizable scorecard to see if your stocks pass the numbers test
• http://www.valueinvestingpro.com Value Investing Pro

So out of the 3 valuation methods you mention, which one you like the most and which one you find the most reliable in terms of performance?

I personally like none of the 3, I use FCF Yield but if I had to, DCF. I just think it’s both art and science and assuming you are using realistic assumptions, you will get a reasonable IV

-Alex

p.s. great job as always
.-= Value Investing Pro´s last blog ..Stocks On The Move 9.20.09 =-.

• Floris

Awesome article. Looking forward to reading Greenwalds book.

I am quite a risk averse investors and in addition to the analysis you perform (which is more extensive than my own), I always ask myself the question what can go wrong? and attempt to extrapolate value from a worst case scenario. This allows me to get a much better feel for the margin of safety of a stock. Although these models are great as a tool for analysis, radical shifts (blacks swans) should never be discounted.

• http://www.oldschoolvalue.com Jae Jun

@ Value Investing Pro (Alex),

DCF is still my favorite. I find it easy to use and doesn’t require as much messing around with the numbers. If the valuation fits, good, if it doesn’t pass. Although I’ve automated the EPV, I’m afraid that I’ll be skipping over all the required digging around that Greenwald emphasizes.

But since it wont be a technique I’ll solely be relying on, it should be fine.

I never really used FCF yield because it misses out on much more opportunities, but whatever works for the individual right?

Thanks for the comment.

@Floris,

That’s a great question to ask yourself. I’m always trying to see what the worst case scenario is myself and then make adjustments accordingly.

e.g. MHH and IGOI I felt were past their worse case scenario. In IGOI’s case, they lost a customer accounting for 40% of their revenue but their stock price didn’t drop a cent. I found that intriguing and after some more reading, felt that the only direction left was up.

120% later, I’m glad I don’t like risk as well

• http://www.stockpursuit.com Stock Pursuit

I just noticed IGOI is in play. I forgot it was a value stock until i remembered you were all over it not that long ago. nice call. What do you think of the big news on the 14th? if you still are following it i guess
.-= Stock Pursuit´s last blog ..More On Watch List =-.

• http://www.oldschoolvalue.com Jae Jun

I still hold it and from my calculations even with a pessimistic view, the lowest it should trade at should be \$1.50.

But with plenty of cash and the company doing a great job of increasing distribution channels, the upside is much more if things start improving.

The news about Verizon was the only thing leading up to this. They’ve been steadily announcing new deals and product launches that should do well for them.

• Peter

Hi Jae Jun,

I reviewed your new spreadsheets. They look exceptional!!! I might have missed it, but did you include the maintenance capex calculation in the new spreadsheet? In addition, For those who have bought the premium package, would you consider some sort of renewal fee once the year has expired instead of buying the entire premium package again.

• http://variantperception.wordpress.com PlanMaestro

That is great analysis Jae. Though it provoked me a cognitive dissonance. What is Microsoft’s competitive advantage

Usually value investors talk about:

1. Brands but that is accumulated marketing
2. Distribution Channels: can be valued too
3. Patents but that is accumulated R&D
4. Location but that is real estate mispriced in the books
etc.

But all those are assets that we can put a value to. So if you are valuing ALL Microsoft’s tangible and intangible assets at replacement cost what is their competitive advantage?

In the competitive advantage literature is usually resources (but you added them all back), culture (that is clearly not Microsoft’s case) or a difficult to copy mix of the resources. Maybe that they can tie up innovations to their OS?
.-= PlanMaestro´s last blog .. Variant Perceptions =-.

• http://www.oldschoolvalue.com Jae Jun

@ Peter,
Yes the maintenance capex calculation is included. It’s towards the bottom section of the EPV page.

As for a renewal type payment model, I’ll have to consider that. Selling spreadsheets isn’t a business model that I thought too much about so the business aspect is a little raw and primitive at the moment. I’ll see what I can come up with but for now, I’ll keep it as a repurchase method. One way I could do is email all the premium buyers from more than 1 year ago and offer a discount or something.

Thanks for the suggestion Peter.

@ PlanMaestro,

While branding is based on marketing, for leaders such as KO, MSFT or even GOOG, I don’t believe they need to market to still be effective.

But the EPV method doesn’t value all of the tangibles and intangibles. It is only a representation of the minimum that is required to start a business. If a competitor even wanted to come close to competing with MSFT, they would have to spend far far more than what is stated as the asset valuation I came up with.

So this method doesn’t really add back everything. Only what a competitor may have to pay today in order to open it’s doors.

One other thing that EPV considers is that a company like Nucor may have started when inflation was at its lowest point and so of its PPE were bought at low prices, but if a new entrant was to try and come into the market, the same piece of equipment, although newer, would probably cost much more than what Nucor paid.

The difference between the EPV price and the asset valuation is what Greenwald calls the value of the franchise. Its moat, which, in the end, is still a subjective topic and one that I know you are well aware of.

• http://www.stockpursuit.com Mark

There’s sort of complex arb play with FTBK. This one is prob a little to complex for me but maybe you or somebody else can figure out the best way to play it…I’m thinking just buy some FTBK maybe if it dives soon…???
.-= Mark´s last blog ..Trading Update =-.

• http://valueinvestortoday.com Jim

@ Value Investing Pro,

How exactly do you find the intrinsic value of a business by using the FCF Yield metric? It really doesn’t tell me anything regarding a margin of safety.
.-= Jim´s last blog ..The Wall Of Shame =-.

• david

I was just wondering why you used 2 when multiplying the growth rate in calculating intrinsic value instead of 1.5 as it is in the graham intrinsic spreadsheet? When do you consider using the multiplier of 2 in calculations? Any thoughts are greatly appreciated as I am just beginning my investment journey.

• http://www.oldschoolvalue.com Jae Jun

Hi David,

Which multiplier of 2 are you referring to? I made the Graham adjustments and published it after I wrote this article btw.

• Alex

Hello Jae,

I’ve also read Greenwald’s book.

What I didn’t get was why didn’t he use FCF/Owners Earnings instead of adjusted Ebit. He seems to make a load of loose estimates, where he could have started with FCF and made the adjustments for the biz cycle etc.

Thanks

• http://www.oldschoolvalue.com Jae Jun

By using FCF numbers, you wouldnt be able to include the other aspects of the business which adds to the earnings power. He makes a valid point that R&D, SG&A and other line items need to be added back. If you did this just FCF, you would either be double counting it or not considering the business as a whole but rather just the profitability.

• Borislav Koev

Jae

Thanks for your post, it’s great but I have some questions:

– In step 2 you “add a certain percentage of SG&A and R&D back to earnings. I prefer to keep things simply by adding back 25% for both.”
What is the logic behind that?

– When calculating Earnings Power Value in step 4 you “add back 20% of D&A as I do”. Why only 20% and not the whole amount?

Thanks!

• http://www.oldschoolvalue.com Jae Jun

The best way would be to read the book but to answer simply

1. EPV is for finding the earnings power. SG&A and R&D add to the future earnings but not 100% of it. So Greenwald suggests you add 25% as a conservative measure.
2. Same reason as above except we are dealing with capex when adding back D&A. If I add back 20%, I am assuming that the remaining 80% will more than cover the maintenance capex.
It’s on page 125 of the book. Last paragraph on the page.

• Borislav Koev

However, something is still bothering me:
you said that 80% of D&A should be enough to cover the maint. CAPEX (I agree), but then the next step in the analysis is to “subtract maintenance capital expenditures and divide by the discount rate.”
This means that you count the maint. CAPEX twice!!!

Thanks again!

• http://www.oldschoolvalue.com Jae Jun

Well 20% of D&A is added back first because I am assuming the 20% of hard assets creates earnings. And remember that D&A is not capex so it never is taken twice. Maintenance capex is calculated completely separately.

• Jason

Hey Jae,

Great post but one thing I don’t get it. In GreenWals’s slide 14. His Net repro value is to subtract all Liab. (A/P, AT, AL, Debt, Deferred Tax, Reserve) from Reproduction Asset Value but your model only subtract non interesting bearing debt and excess cash.

Is any assumption behind that?

Thanks

Jason

• Rayane

Hi Jae,

Thank you for your great site.

I am a bit puzzled by your calculation of net reproduction asset value.

I don’t think you should substract non interest bearing debt AND the cash not required to run the business from Asset Reproduction Value (ARV). I think that by doing so, there are chances that you could underestimate the capital (equity and interest bearing debt) needed to reproduce the business.

In my opinion, you should indeed substract non interest bearing debt from ARV, and then cash not required to run the business IN EXCESS of non interest bearing debt and 1-2% cash required to run the business. That way, you find the true ARV, which you can compare to the EPV.

I understand from Greenwald’s book that there are several ways to use Net AVR. One in comparison with the EPV, which is the calculation I just described, from my understanding. Second, by comparing the same calculation with the market value of the company, as measured by Market Capitalization + Market Value of Debt – Cash. Third, by calculating what I would call a net net AVR : AVR – All reproduction liabilities, which you would compare to the market cap of the company.

What do you think ?

Best regards,

Rayane

• Greg

Very interesting stuff however, and this is fairly off the cuff not having read Greenwald’s stuff, what about the basic common sense fact that you can’t reproduce Microsoft? Even if you had the money, which is questionable, you could not today go out and buy the worldwide market share that Windows OS, Office, Server, XBOX have. Can’t do it from scratch. You can’t just go out and reproduce all the patent’s, processes, etc.

• http://www.oldschoolvalue.com Jae Jun

@ Greg,
But that’s exactly what Greenwald and I’ve shown through the EPV. Their franchise value is amazing and eclipses the reproduction value. That difference in EPV and reproduction value is what distringuishes the earnings power and franchise value.

• PSDFinancier

Hey Jae,

Quick question for you on the EPV Calc. I went through Greenwald’s book, and I didn’t see him adding back SG&A and R&D to EBIT in his walkthrough in the book. Nor do I see that adjustment made in his lecture slides on calculating EPV. I know he makes an adjustment for SG&A and R&D in calculating his asset reproduction value. I was wondering if you could explain why you make the adjustment in your EV Calc, and whether perhaps Prof. Greenwald discusses this somewhere that I am missing. Thanks!

• http://www.oldschoolvalue.com Jae Jun

Hi PSD,
Here’s the quick answer from the book pg 124-125

To get a more accurate picture of the earnings power with zero growth, we ought to add some of the R&D expense back to operating earnings…Fourth, the adjustment for R&D can also be applied to SG&A…. we will add back 25% of the total.

• http://www.valueinvestingindiareport.com/ Ankur

Jae,

Thanks for the excellent example of an EPV analysis. After reading Greenwald’s book, it’s clear that he dislikes using DCF analysis because it requires too many assumptions. But it’s interesting that his own methodology for calculating reproduction asset value relies on a significant amount of assumptions particularly for calculating goodwill. Also, to answer PSD’s question the reason you add back R&D and SG&A expense is that EPV assumes a zero growth scenario. Greenwald basically assumes that 25% of annual SG&A and R&D expense is associated with growth. Thus, in an EPV no growth scenario 25% of both expense lines should be added back to EBIT.

• http://www.oldschoolvalue.com Jae Jun

Hi Ankur,

I would argue that Greenwald doesnt like DCF because it requires forward looking assumptions while his assumptions on reproducing book value and goodwill are present based.

• Alex

Hi Jae,

I am reviewing your EPV valuation and I think what you are doing is amazing.

When looking at step 1: adjusting balancesheet, I notice something strange. At the book value, I come up with the total assets of: 77.888,0 instead of 80.101,0.

Here’s the calculation:

Cash & Equivalents: 31.447
Receivables: 11.192
Inventories 717
Current defered income taxes: 2.213
Other Current Assets: 3.711

Total current assets: 49.280.

Intangibles: 14.262
Property/Plant etc..: 7.535
LT investments: 4.933
Other long term assets, total: 1.878

Total assets: 77.888

Secondly the reproduction value should be a bit higher. On the reproduction asset value you only made adjustment in the goodwill, which is 50% off and equals to 6.251.

So the end result should be: 80.101 – 6.251 = 73.850 instead of 68.916,5

Pls correct me if I am wrong,

Alex

• http://www.oldschoolvalue.com/ Old School Value

Remember that with reproduction value, you are not necessarily adding
the full value. Depending on the company this method requires adding
back portions of it if necessary. One example is that I didn’t add back
all of goodwill.

Right now, I didn’t go through the numbers in detail because I wrote
this 3 years ago so the numbers are hazy, but the concepts are exactly
the same. Just different numbers maybe.

• alex

I am not so familiar with EPS, so I am trying to figuring out, but thanks for the quick answer!

• http://www.facebook.com/mike.mcdonagh.5 Mike McDonagh

Alex – Why did you input \$3,711 for Other Current Assets? It should be \$5,924. That is the difference between your total asset value of \$77,888 and the correct total asset of \$80,101.

To your second point. You are correct, the difference between your Reproduction Asset Value (RAV) of \$73,849.5 and Jae’s RAV of \$68,916.5 is that Jae forgot to add in Long Term investments of \$4,933 because that is the exact difference.

• http://www.facebook.com/mike.mcdonagh.5 Mike McDonagh

You must have subtracted Current Deferred Income Taxes, but you really should of added that figure

Hope my posts have helped.